Chinese mitten crab (Eriocheir sinensis) is one of the most favorite seafoods in the Asian areas. The morphologies of its carapace exhibit pronounced intraspecific variation, making them highly informative phenotypic traits for individual identification, geographic origin tracing, and anticounterfeiting management in aquaculture and seafood quality control. Accurate localization of the key landmarks on the carapace is one of the most critical procedures for the quantitative phenotype analysis, precise individual recognition, and processing tasks. Conventional approaches can rely predominantly on the manual visual assessment and expert judgment, unsuitable for the large-scale and automated applications in modern aquaculture, due to the time-consuming, labor-intensive, and prone to inconsistency. In this study, a high-precision framework of the keypoint detection (named YOLO-FMC-pose) was specifically designed for the carapace of the Chinese mitten crab. A dataset was constructed with the high-resolution images of the crabs. Multiple geographic origins were selected from the Liangzi Lake, Junshan Lake, and Yangcheng Lake. Then, the 35 representative landmarks on the carapace were selected and manually annotated for the biological interpretability and structural completeness. Data augmentation was applied to improve the robustness and generalization of the model, including the random rotation, scaling, brightness and contrast adjustments, and horizontal flipping. Diverse real-world conditions were simulated during imaging. The YOLO-FMC-pose model was based on the lightweight YOLO11n-pose backbone. Three improvements were also incorporated to enhance the frequency sensitivity, multi-scale semantic integration, and attention-guided spatial representation. Firstly, a C3K2FD module was integrated with the Frequency Dynamic Convolution (FDConv). Rich frequency-dependent features were captured in response to the high-frequency edge details and low-frequency smooth textures in the carapace. Secondly, a Mixed Aggregation Network (MANet) was incorporated in the Neck stage. Multi-scale features were aggregated to distinguish the subtle structural differences among landmarks. Thirdly, the Convolutional Block Attention Module (CBAM) was integrated into the detection head. Both channel and spatial attention mechanisms were employed to emphasize the informative regions while suppressing irrelevant background noise. Three modules functioned synergistically to accurately capture the spatial arrangement and fine-grained structure of the critical landmarks. Extensive experiments were conducted to evaluate the performance of the YOLO-FMC-pose against several state-of-the-art lightweight detection models, including the YOLOv8n-pose, YOLOv10n-pose, YOLOv12n-pose, and the original YOLO11n-pose. The results demonstrated that the YOLO-FMC-pose achieved superior performance over multiple metrics. Specifically, the better performance was achieved in a precision of 97.98%, a recall of 97.00%, a mAP_0.5 of 98.27%, and a mAP_0.5:0.95 of 73.28%. Compared with the original YOLO11n-pose, these values represented the absolute improvements of 3.33, 2.33, 2.94, and 13.08 percentage point, respectively. The normalized mean error (NME) of the predicted keypoints was reduced to 3.835%, indicating the highly accurate spatial correspondence between predicted and ground-truth landmarks. The inference speed remained at 7.5 milliseconds per image, indicating its feasibility for real-time deployment in the aquaculture processing and quality control pipelines. Attention heatmaps revealed that the YOLO-FMC-pose was consistently focused on structurally significant regions of the carapace, including edges, protrusions, and concavities, whether the imaging device or lighting conditions. The high robustness and reliability of the model were obtained to identify the critical anatomical features during diverse acquisitions. The YOLO-FMC-pose was provided for the precise keypoint detection in the downstream applications, such as individual crab identification, geographic origin verification, anti-counterfeiting labeling, and traceability systems. In summary, an effective approach was presented to extract the fine-grained phenotypic features in the Chinese mitten crab. Multi-module deep learning strategies were integrated for high accuracy, robustness, and efficiency. Landmark detection of the crab carapace can provide a scalable framework for intelligent aquaculture and aquatic product traceability. Dataset diversity can be expanded under varying environmental conditions. The model was also deployed on the edge and embedded devices for real-time applications. Keypoint detection was integrated with the multi-dimensional phenotypic analysis for individual identification and quality assessment.

Agricultural product quality and safety traceability can effectively enhance the trust between supply chain entities and consumers. It is often required for precise quality and safety recall mechanisms to ensure national food safety in recent years. This article aims to systematically summarize the basic concepts and classifications of agricultural product quality and safety traceability in China. The history of traceability was elaborated in three stages: institutional frameworks, platform construction, and digital transformation. The key technologies were successfully integrated, such as the Internet of Things, big data, blockchain, and artificial intelligence. Significant progress was achieved in information perception, data processing, anti-counterfeiting traceability, and intelligent analysis. Its technological empowerment was elucidated across the three-dimensional layer, including the information perception, processing, and decision-making layer. The advantages and disadvantages of these traceability applications were discussed to summarize the existing traceability platforms, national standards, industry standards, and local standards. However, some challenges remained in the data sharing and integration, including the severe data silos, diverse traceability models, as well as the less standards and specifications. Additionally, the high costs and the limited integration of emerging technologies with the traceability framework have restricted the promotion and application of such systems. The traceability technology was also aligned with the market-oriented applications and platform implementation in practice. The optimal systems were gradually improved the standardization frameworks. Future research and application can focus on the following aspects. In the traceability information perception, the intelligent equipment (such as embodied intelligence and low-altitude drones) will play a significant role in the logistics and distribution. Hardware development can also drive toward greater intelligence and automation. In traceability information processing, large-scale models and quantum blockchains can be explored in data processing and intelligent decision-making. In the traceability information interaction, cutting-edge technologies can be applied, like big data, the Internet, and the Internet of Things. Furthermore, the next-generation technologies were integrated with the traceability system, such as 3D printing, the metaverse, and digital twins. The traceability standards can evolve into cross-platform, cross-regional, cross-departmental, and even cross-border collaborative traceability. Application-oriented standard leadership can be strengthened for a unified technical standard system. A traceability standard system can be developed to cover the entire supply chain of agricultural products. Finally, the findings can provide theoretical support and practical guidance to advance the intelligent and collaborative system.

As the global food safety problem has increasingly become severe and supply chain reforms and food recalls are costly and challenging, it is particularly important to establish agricultural food traceability systems. Countries around the world are paying increasing attention to developing food safety traceability systems. Under the background of smart agriculture, information and communication technology can be further improved through blockchain infrastructure to achieve new farms and digital agriculture, thus ensuring end-to-end safety traceability of agricultural products from planting to sale. Through the analysis of the latest domestic and international research, this paper systematically elaborates on recent progress in research on agricultural food traceability based on blockchain technology. It dissects the processes of agricultural food supply chains, and proposes a basic architecture of blockchain in the field of agricultural food traceability. In addition, this paper summarizes the application of blockchain in agricultural food traceability, focusing mainly on blockchain combination with cloud-edge computing, encryption technology, storage optimization, consensus mechanism improvement, and smart contract design. It also points out the challenges in data security, storage scalability, regulatory difficulties, and practical applications. Finally, it proposes future direction for blockchain technology in agricultural food traceability such as strengthening security supervision, improving blockchain scalability, and empowering food traceability with emerging technologies. It emphasizes the opportunities and challenges in the application of blockchain technology in the agricultural food supply chains.

A blockchain-based agri-food traceability can be expected to remedy the inherent trust within agri-food supply chains. However, the immutable and redundant nature of traditional blockchain storage has posed formidable challenges, particularly on the scalability of data storage for blockchain nodes. Agri-food traceability has also been limited to historically impede the widespread adoption in recent years. In this study, an efficient storage and query model was introduced to specifically design for the agri-food supply chain traceability, in order to leverage the concept of a redactable blockchain. The cyclical nature of agri-food traceability data was examined to form the model. The lifecycle of the supply chain was analyzed for the agri-food products. The opportunities were identified for streamline data management. A key innovation involved the strategic offloading of traceability data over the lifecycle. Storage resources were optimized to preserve the essential traceability functionalities. The information remained accessible throughout the supply chain journey. Furthermore, the counter Bloom filter was incorporated to enhance the operational efficiency of data offloading. The high false positive rates were reduced for the data manipulation in redactable blockchains. False positives were effectively reduced to significantly enhance the overall query efficiency of the agri-food traceability system, thereby facilitating expedient access to accurate supply chain information. A counter Bloom filter was operated to leverage the probabilistic hashing techniques. The large datasets were efficiently managed and queried to minimize the occurrence of false positives. A robust mechanism was translated to verify the accuracy of traceability information post-data offloading in the context of the agri-food traceability system. A compact representation of recently offloaded data was maintained to employ the efficient hash functions. The counter Bloom filter effectively reduced the likelihood of mistakenly, in order to identify non-existent data during queries. The efficacy of the improved model was underscored to validate the comprehensive empirical data. The extensive experiment was conducted over a simulated 60-month operational period. Notably, the better performance of the model was achieved with a remarkable 48.70% reduction in storage volume, compared with the conventional agri-food blockchain traceability systems. This reduction was attributed to the strategic lifecycle-based data management, and the storage was optimized without compromising data integrity. The improved model was often verified in a simulated scenario involving 1 000 block records and a 30% data offloading rate. There was a notable 21 percentage points decrease in the false positive rates, indicating the efficacy of the integrated counter Bloom filter with the data accuracy post-offloading. Moreover, there was a commendable 19.02% enhancement in the data query efficiency, compared with the traditional approaches. The compelling solution fully met the operational demands of large-scale agri-food supply chain environments. The blockchain-based agri-food traceability was presented to facilitate the widespread deployment. A significant advancement was achieved in data integrity with the pragmatic storage and query optimizations in the field. Beyond technical innovation, a robust framework was offered to enhance transparency, accountability, and consumer confidence across agri-food supply chains. Looking ahead, the scalability and adaptability of the improved model can promise to support the diverse applications within the agri-food sector. Product authenticity and quality assurance can be enhanced to enable efficient recalls and regulatory compliance. As the blockchain continues to evolve, insights can be gained to pave the way for future advancements in agri-food traceability and industry-wide transformation toward a more resilient and sustainable supply chain.